Ting et al. Trials (2018) 19:310 https://doi.org/10.1186/s13063-018-2649-9

STUDYPROTOCOL Open Access Effectiveness and sustainability of a structured group-based educational program (MEDIHEALTH) in improving medication adherence among Malay patients with underlying type 2 diabetes mellitus in State of : study protocol of a randomized controlled trial Chuo Yew Ting1, Shahren Ahmad Zaidi Adruce1*, Mohamed Azmi Hassali2, Hiram Ting3, Chien Joo Lim4, Rachel Sing-Kiat Ting5, Abu Hassan Alshaari Abd Jabar6, Nor Anizah Osman7, Izzul Syazwan Shuib8, Shing Chyi Loo8, Sui Theng Sim8, Su Ee Lim7 and Donald E. Morisky9

Abstract Background: Amidst the high disease burden, non-adherence to medications among patients with type 2 diabetes mellitus (T2DM) has been reported to be common and devastating. Sarawak Pharmaceutical Services Division has formulated a pharmacist-led, multiple-theoretical-grounding, culturally sensitive and structured group-based program, namely “Know Your Medicine – Take if for Health” (MEDIHEALTH), to improve medication adherence among Malay patients with T2DM. However, to date, little is known about the effectiveness and sustainability of the Program. Methods/design: This is a prospective, parallel-design, two-treatment-group randomized controlled trial to evaluate the effectiveness and sustainability of MEDIHEALTH in improving medication adherence. Malay patients who have underlying T2DM, who obtain medication therapy at Petra Jaya Health Clinic and Health Clinic, and who have a moderate to low adherence level (8-item Morisky Medication Adherence Scale, Malaysian specific, score <6) were randomly assigned to the treatment group (MEDIHEALTH) or the control group. The primary outcome of this study is medication adherence level at baseline and 1, 3, 6 and 12 months post-intervention. The secondary outcomes are attitude, subjective norms, perceived behavioural control, intention and knowledge related to medication adherence measured at baseline and 1, 6 and 12 months post-intervention. The effectiveness and sustainability of the Program will be triangulated by findings from semi-structured interviews with five selected participants conducted 1 month after the intervention and in-depth interviews with two main facilitators and two managerial officers in charge of the Program 12 months after the intervention. Statistical analyses of quantitative data were conducted using SPSS version 22 and Stata version 14. Thematic analysis for qualitative data were conducted with the assistance of ATLAS.ti 8. (Continued on next page)

* Correspondence: [email protected]; [email protected] 1Institute of Borneo Studies, Universiti Malaysia Sarawak, Kota Samarahan, Sarawak, Malaysia Full list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ting et al. Trials (2018) 19:310 Page 2 of 13

(Continued from previous page) Discussion: This study provides evidence on the effectiveness and sustainability of a structured group-based educational program that employs multiple theoretical grounding and a culturally sensitive approach in promoting medication adherence among with underlying T2DM. Both the quantitative and qualitative findings of this study could assist in the future development of the Program. Trial registration: National Medical Research Register, NMRR-17-925-35875 (IIR). Registered on 19 May 2017. ClinicalTrials.gov, NCT03228706. Registered on 25 July 2017. Keywords: Culturally sensitive, Type 2 diabetes mellitus, Medication adherence, Malay patients, Group-based educational program, Randomized controlled trial

Background significantly associated with better glycaemic control, Diabetes is an epidemic chronic disease caused either by less hospital visits and admissions, and lower medical costs. the failure of the pancreas in producing sufficient insulin On the other hand, a lower adherence rate was significantly or by the failure of the body cells in utilizing the insulin associated with poor medication tolerance, frequency of produced by the pancreas effectively [1]. The majority medication intake (more than two times a day), having (95%) of patients with diabetes have type 2 diabetes concomitant depression and negative belief about the mellitus (T2DM), which is more common in adults than medications. Consequently, patients who poorly adhere to in children [1, 2]. A national health and morbidity medications take more medications owing to poor gly- survey (NHMS) carried out by the Ministry of Health, caemic control and the development of micro- and macro- Malaysia (MOH), in 2015 revealed that 17.5% (95% Con- vascular complications [11]. Furthermore, the condition fidence interval 16.6, 18.3) of Malaysians were found to further worsens their adherence owing to more complex have diabetes [3]. In comparison with the results of the medications and more side effects experienced [7]. This first NHMS and the second NHMS conducted in 1986 inevitably increases the financial burden and wastage to and 1996, respectively, the recent findings again showed health services [12]. Hence, breaking the vicious cycle that there is a steady increase in the prevalence of dia- should be an urgent priority for all stakeholders. betes among Malaysians [4]. Besides, the national survey In Malaysia, the Pharmaceutical Services Division (PSD) found that there is a significant difference between differ- of the MOH launched in 2007 a campaign called “Know ent ethnic groups. Indians were found to have the highest Your Medicine” (KYM) to promote the quality use of prevalence of diabetes (22.1%, 95 CI 19.2-25.3), followed medicines [13]. The campaign utilizes mass media, social by the Malays (14.6%, 95 CI 13.8-15.5), the Chinese (12. media and a group-based educational program (GBEP) to 0%, 95 CI 10.7-13.5) and last the other indigenous groups deliver the message to the Malaysian public. The messages (10.7%, 95 CI 8.8-13.0). Even though Indians were found conveyed include information on their medication man- to have a greater percentage of T2DM than Malays, agement, such as why, how and when to take medicines, Malays have a higher total number of T2DM patients reporting adverse drug events, awareness on the rational owing to their bigger population size in Malaysia than use of medicines and medications that need special precau- Indians, which is the third largest ethnic group in tions. In particular, assuring and improving medication Malaysia [5]. adherence among patients is one of the important mes- Amidst the high disease burden, non-adherence to med- sages conveyed through the campaign. Moreover, the PSD ications among patients with T2DM has been reported to of Sarawak State Health Department further expanded the be common and devastating [6]. It has been estimated that scope of the GBEP by formulating a 3-h pharmacist-led more than 50% patients fail to achieve recommended culturally sensitive structured GBEP in promoting medica- glycaemic goals due to non-adherence to diabetic medica- tion adherence among T2DM patients [14]. The official tions [7, 8]. There is no exception for Malaysia as a re- name of the structured GBEP is “Know Your Medicine – cent national survey revealed that 73.1% of Malaysians Take it for Health” (MEDIHEALTH). Notably, MEDI- who are on medication did not adhere to the medica- HEALTH is a culturally sensitive and culturally appropriate tions prescribed [9]. program that is tailor-made to suit the cultural differences A recent systematic review [10] summarized that the of the major ethnicities in the State, including Iban, factors that contribute to poor adherence among T2DM , Malay and Chinese. In addition, MEDIHEALTH patients are age, ethnicity, health beliefs, medication cost, was specifically designed to complement the individual ap- insulin use, health literacy, medication cost, co-pays, med- proach in improving medication adherence among T2DM ical insurance coverage and primary non-adherence. Fur- patients called Diabetes Medication Therapy Adherence thermore,ontheonehand,agreateradherenceratewas Clinic (DMTAC) initiated by the PSD, MOH [15]. Previous Ting et al. Trials (2018) 19:310 Page 3 of 13

literature supports the employment of a structured GBEP attitude towards the behaviour, perceived subjective so- such as MEDIHEALTH because it has numerous irreplace- cial norms towards the behaviour and perceived behav- able benefits, including (1) validation, (2) normalization of ioural control towards the behaviour and is mediated by experience, (3) reduction of isolation, (4) sense of belong- the intention to act. Attitude towards the behaviour is ing and (5) enhanced self-esteem [16]. The design and how one evaluates the advantages or disadvantages of content of MEDIHEALTH will be further elaborated in the performing a behaviour. Subjective social norm is the “Methods/design” section. social expectations of a behaviour perceived by an indi- While the previous literature has shown that a GBEP in vidual. Perceived behavioural control is personal percep- promoting self-management among T2DM patients has tion about the difficulty and capability of performing a proven to be effective [17], little has been done to examine behaviour. Ajzen further revealed that perceived behav- the effectiveness of a structured GBEP in improving medi- ioural control has a direct effect on the behaviour when cation adherence among T2DM patients [18]. Further- a given behaviour has less volitional control. Perceived more, we echo what has been mentioned by Edmondson behaviour control could reflect both external (such as and colleagues in that “very few” studies of behaviour time and money) and internal (such as skill and informa- change interventions had actually tested the mechanism tion) factors, which is similar to the self-efficacy con- of behavioural interventions [19]. Such a gap in the know- struct conceptualized by Bandura [27]. Thus, the level of ledge has caused researchers to remain unclear on why perceived behavioural control of an individual will deter- and how an intervention works or fails and thus has mine the persistence of a given behaviour over time in limited future intervention development [19]. Hence, by the face of obstacles and setbacks. For a better illustra- investigating the effectiveness of the structured GBEP tion of the value of the TPB in explaining behaviour in MEDIHEALTH and its mechanism, this study intends to this study, medication-taking behaviour (medication ad- provide managerial implications to the PSD, Malaysia, and herence) is influenced by attitude, perceived social norm to bridge the gap in the extant body of knowledge. and perceived behavioural control over medication ad- Apart from focusing on the effectiveness of MEDI- herence, and all three factors are mediated by intention HEALTH, its sustainability is also an important component towards medication adherence, whereas perceived be- to the organizations that formulate, implement and finance havioural control has also a direct effect on the behav- the program [20, 21]. Moreover, health interventions that iour. A recent systematic review and meta-analysis on are not sustainable will result in participants becoming the use of TPB in the studies of adherence to treatment disillusioned and hinder participants from future engage- in chronic illness further supports the feasibility of the ment in any health interventions that could be beneficial to TPB in the current study [24]. The TPB was found to ac- them [21]. Thus, this study will also evaluate the sustain- count for 32.92% of the variance in intention and 9.18% ability of the program since it is still at the early phase of of the variance in behaviour, and all the relationships implementation. Notably, researchers employ the perspec- between variables were found to be consistent with the tives of Pluye and colleagues in conceptualizing sustainabil- original hypotheses of the TPB. ity, which involves organizational routines and institutional On the other hand, the IMB is a simple and the latest standards [22].Moreover,weemployaqualitativeap- behavioural model that has a high predictive value on proach to evaluate the sustainability of the program “to long-term medication adherence [28]. It had been studied understand the phenomenon, refine hypotheses, and de- among patients affected by tuberculosis [28], HIV [29]and velop strategies to promote sustainment”, as recommended T2DM [25], who are required to adhere to long-term by Stirman and colleagues [23]. medication to achieve a good clinical outcome. The IMB In terms of theoretical grounding to evaluate the change posits that to achieve behavioural change, one should have in medication-adhering behaviour, on the one hand, the adequate behaviour-related information; health behaviour theory of planned behaviour (TPB) was chosen as the motivation, which consists of personal motivation and sub- underlying theory in evaluating the effectiveness of MEDI- jective motivation; and necessary skills to perform a spe- HEALTH as it has the strongest empirical evidence in cific health behaviour [29]. Notably, there are similarities medication adherence studies [24]. On the other hand, the between IMB variables and TPB variables. Firstly, the information-motivation-behavioural skills model (IMB) motivation constructs of the IMB, which consist of per- was chosen as the supporting theory as it has been sonal motivation and social motivation, are like the attitude empirically tested on medication adherence among T2DM and subjective norm of the TPB, respectively. Secondly, the patients [25]. behavioural skill variable is like perceived behavioural On the one hand, the TPB is a theory founded in the control. In terms of the relationship between the three field of social psychology and is best suited to describe independent variables, the behavioural skill variable serves volitional behaviours, especially health-related behav- as a mediator that mediates the effect of behaviour-related iours [26]. It posits that a behaviour is influenced by information and health behaviour motivation towards the Ting et al. Trials (2018) 19:310 Page 4 of 13

health behaviour. However, both behaviour-related infor- H1a: Increase in intention to adhere would be mediated mation and health behaviour motivation have a direct by improvements in attitude towards adherence after effect on health behaviour. baseline, which is achieved by participating in By comparing the constructs of the IMB and TPB in MEDIHEALTH. explaining a behaviour, regardless of the relationship H1b: Increase in intention to adhere would be between constructs, the main difference is that the mediated by improvements in subjective norm towards IMB has an additional construct, which is behaviour- adherence after baseline, which is achieved by related information. In the IMB, the behaviour-related participating in MEDIHEALTH. information construct has a direct effect on health be- H1c: Increase in medication adherence would be haviour and behavioural skills. By noting the similarity mediated by improvements in intention to adhere after of both theories, the researchers hypothesize that the baseline, which is achieved by participating in behaviour-related information construct of the IMB MEDIHEALTH. serves as an independent variable which has a direct H1d: Increase in intention to adhere would be effect on the perceived behavioural control construct mediated by improvements in perceived behavioural and the medication adherence construct of the TPB. control towards adherence after baseline, which is Such a hypothesis is aimed at increasing the explanatory achieved by participating in MEDIHEALTH. power of the underlying theory towards medication-taking H1e: Increase in intention to adhere would be mediated behaviour through extending the TPB. Nevertheless, by improvements in adherence information after MEDIHEALTH is included as one of the independent baseline, which is achieved by participating in variables that will improve the scoring of extended TPB MEDIHEALTH. variables. The conceptual framework of this study is H1f: Before the intervention, there are no significant presented in Fig. 1. differences of medication adherence level and the This study aims to investigate the effectiveness of MEDI- psychosocial variables related to it among the HEALTH in improving medication adherence among participants between the intervention group and the Malay communities with underlying T2DM in the Sarawak control group. State of Malaysia. The specific objectives of this study are H1g:After1,3,6and12monthsoftheprogram, (1) to measure the effectiveness of MEDIHEALTH in im- the medication adherence levels among the proving the medication adherence level and the compo- participants in the intervention group are nent of the extended TPB, (2) to identify the component of significantly greater than the medication adherence the extended TPB that predicts medication adherence after levels before the intervention. participating in MEDIHEALTH, and (3) to investigate the H1h: After 1, 3, 6 and 12 months of the program, the sustainability of the program. Based on the first specific ob- medication adherence levels among the participants in jective, eight hypotheses are to be tested: the intervention group are significantly greater than the

Fig. 1 Conceptual framework of the study (TPB constructs are in black while IMB construct is in green) Ting et al. Trials (2018) 19:310 Page 5 of 13

medication adherence levels of the participants in the the Petra Jaya Health Clinic (PJHC) and the Kota control group. Samarahan Health Clinic (KSHC). Notably, Malays who are diagnosed with T2DM receive continuous consult- For the second specific objective, seven hypotheses are ation and monitoring at the clinics. Hence, with the to be tested: aim of recruiting suitable respondents that could yield highly generalizable data to the studied population, H2a: Improvement in attitude towards adherence will PJHC and KSHC have breen chosen as the study sites contribute to the increase in intention to adhere. to recruit respondents and collect data. Moreover, H2b: Improvement in subjective norm towards PJHC and KSHC are to be included as the KYM cam- adherence will contribute to the increase in intention paign sites. Thus, the researchers incorporated this to adhere. study into the KYM campaign by obtaining prior ap- H2c: Improvement in perceived behavioural control proval from Sarawak Pharmaceutical Services Division. towards adherence will contribute to the increase in intention to adhere. Participants H2d: Improvement in perceived behavioural control Sample size towards adherence will contribute to the increase in The current study is the first to investigate the effectiveness medication adherence. and sustainability of the program since the program was H2e: Improvement in intention to adhere will developed in late 2016 and began to test run in early 2017. contribute to the increase in medication adherence. Hence, we aim to achieve such objectives with minimum H2f: Improvement in adherence information will resources and time required before this program is ex- contribute to the increase in perceived behavioural panded to other Malay communities in Malaysia. This is a control towards adherence. study of a continuous response variable from independent H2g: Improvement in adherence information will control and experimental subjects where one control per contribute to the increase in medication adherence. experimental subject is planned. In a previous study [30], the response within each subject group was normally dis- Methods/design tributed with a standard deviation of 1.8. If the true differ- Study design ence in the experimental and control means is 1, we need Given the nature of the research problem, an experimental to study 69 experimental subjects and 69 control subjects studydesignwillbeutilizedtoexaminetheeffectivenessof to be able to reject the null hypothesis that the population the structured GBEP MEDIHEALTH. In particular, it was means of the experimental and control groups are equal a prospective, multicentre and parallel-design randomized with a probability (power) of 0.9 [31]. By estimating 30% of controlled with two treatment groups. The protocol is writ- dropout or incomplete data, a minimum sample size of ten in accordance with the Standard Protocol Items: Rec- 180 with 90 for both groups is pre-determined. ommendations for Interventional Trials (SPIRIT) checklist (Additional file 1), and its figures are illustrated in Fig. 2. Selection criteria This trial protocol is registered with ClinicalTrials.gov The inclusion criteria for the participants are (1) (NCT03228706). 8-item Morisky Medication Adherence Scale (MMAS- 8), Malaysian specific, score <6 and (2) Malay T2DM Study population and setting patients; the exclusion criteria are (1) pregnant women, As mentioned above, Malays have the greatest number of (2) patients below 18 years old, (3) patients who have patients with T2DM in the nation. Thus, the researchers severe and enduring mental health problems, (4) pa- made an informed decision by selecting Malay patients tients who cannot listen or read owing to inherited dis- with underlying T2DM as the studied population for the abilities or malfunction, (5) patients who are unable to current study. Moreover, since MEDIHEALTH is a struc- communicate in the Malay language, (6) patients who are tured GBEP formulated by the Sarawak State of Malaysia participating in other studies, (7) patients who decline and has only been implemented in the State, the studied consent to participate and (8) hospitalized patients. population will be entirely Malay patients with underlying T2DM who obtain their medications in Sarawak during Confounding variables the period of this study. Samples were selected from the Based on evidence from previous literature [32, 33], the Malay communities who reside at North City confounding variables included in this study are (1) Hall and Samarahan Division of Sarawak State, which route of administration (oral only or oral and insulin in- have the largest proportion of Malay communities resid- jection); (2) number of medications (one or more than ing in the State [5]. The majority of Malay communities one); (3) frequency of medications (a once-daily dose or are able to access quality healthcare services provided by more frequent); (4) age; (5) sex; (6) highest education Ting et al. Trials (2018) 19:310 Page 6 of 13

Fig. 2 SPIRIT figure. Schedule of enrolment, interventions and assessments. MMAS-8 8-item Morisky Medication Adherence Scale, Malaysian specific; TPB theory of planned behaviour level; (7) monthly household income; (8) employment asked to choose a date to attend the program. After they status; (9) having complications; (10) taking traditional choose their preferred date, their name will be recorded in complementary and alternative medicines; (11) residen- the “List of Participants” with a specific code assigned. tial area (urban or rural); (12) living conditions (having Notably, patients will not be informed of their group social support or living alone); (13) having received dia- assignment and thus will not be not aware of the differ- betic education by a diabetic nurse and (14) enrollment ences between the intervention group and the control with DMTAC. group. Besides, none of the recruiters, who are pharma- cist staff at both health clinics, is aware of subsequent Recruitment treatment allocation throughout the recruitment stage, Prospective and eligible respondents will be recruited to ensure allocation concealment. All research materials consecutively from the two selected health clinics during that contain patients’ information will be coded and their routine scheduled visits before intervention. The kept by the principal investigator to maintain the confi- MMAS-8, Malaysian specific, which has been validated dentiality of respondents. among Malaysians with T2DM [34, 35], will be adminis- tered to obtain the adherence score, enabling us to evalu- Randomization and blinding ate the eligibility of the prospective respondents. Patients The Sarawak Research Society, which is independent who have a low adherence (MMAS-8 score <6) and fulfil from the study team, will be appointed to carry out sim- all the selection criteria will be provided with a study ple randomization. The List of Participants will be handed information sheet and informed consent. Those who agree over to the chairman of the Society to carry out the to participate will be informed about the program, and randomization. The randomization will be conducted using Ting et al. Trials (2018) 19:310 Page 7 of 13

an online randomization program available at http://www. in Malaysia was conducted and summarized [40–43]. Such graphpad.com/quickcalcs/index.cfm, as recommended by review provides valuable insights for the formulation of an Suresh [36]. After that, the participant list, with their code intervention that could tackle the root cause of medication and group assignment, will be kept by the chairman of the non-adherence. In addition, four T2DM Malay patients Society without informing any of the researchers, facilita- were approached to explore their related knowledge, atti- tors or respondents, to ensure the blinding of the three tude, subjective norms and perceived behavioural control parties to treatment allocation. On the day of the interven- towards medication adherence. Four of them were also tion, the chairman and the authorized committee member invited to review the module and had participated in the of the society will register the attendance of respondents simulation of the Program. Feedback was obtained to ensure and inform them about the actual venue of the program. that it fulfils the three components of cultural sensitivity as PJHC and KSHC will have two venues each prepared and mentioned above. Nonetheless, as the studied population is available during the period of the study, with one venue Malay patients with underlying T2DM who reside in allocated to the intervention group and the other one allo- Sarawak,thecontentofthemodulewillbeinthe cated to the control group. None of the participants will be Malaylanguageandallfacilitators will communicate aware of the difference between the venues assigned to with participants using Sarawak’s Malay dialect. As the them, whereas the facilitators will be. However, participants dialect has further derived into different slangs geo- will know whether they are assigned to the control group graphically [44], such as Kuching, Samarahan, Betong as there will not be any form of information provided to and , the MEDIHEALTH program will be orga- the participants in this group. Hence, the blinding of partic- nized for the Malay communities who reside in Petra ipants and facilitators will end at the intervention stage. Jaya () and Kota Samarahan (Samarahan The blinding of the researchers will be continued towards Division) separately. The structured GBEP is conducted the publication stage. During the post-intervention follow- only once for 3 h and facilitated by one main facilitator up, the researchers will still be blinded to the allocation of and three assistant facilitators. participants, as the participant list with group allocation will be kept by the chairman of the society. Facilitators Facilitators who have experience in conducting the inter- Intervention design vention at least five times will be eligible as facilitators for Intervention mapping this study. Apart from training the facilitators to be familiar As mentioned before, the structured GBEP MEDIHEALTH with and consistent in conducting the intervention, their employed both psychosocial and behavioural approaches, training will also include (1) learning how to eliminate con- with multiple behavioural theories as theoretical grounding. cerns and questions the patients come across pertinent to It was mapped based on the application of behaviour medication adherence with the mnemonic tool “ADHERE” change theories as recommended by Slater [37]. Moreover, adopted from Soto-Greene and colleagues [45]; and (2) it is a culturally sensitive structured GBEP that em- learning patient-centred communication as advocated by ploys three components of culturally sensitive inter- Jones [46]. It encourages facilitators to be active listeners vention, including bilingual facilitators and materials, and learn how to encourage the participants to think about literacy-appropriate materials and social support [38]. and understand the problems they face with medication The intervention mapping and content of the inter- adherence and help them decide to act. It also involves vention is available in Additional file 2.Notably,this non-verbal communication skills such as nodding and stage-of-change framework, which incorporates the making eye contact to show genuine interest in and social learning theory, elaboration likelihood model, concern about the issues raised by the participants as the theory of planned behaviour, self-efficacy theory and intervention embraces the philosophy of patient empower- information-motivation-behavioural skills model, was ment, which had been found to be effective in engaging used to map the program but does not serve as an em- patients with diabetes to produce behavioural change [47, pirical measure of the mechanism. The empirical mea- 48]. Hence, the facilitators will be trained to employ a non- sures of the mechanism in our study are the variables didactic approach in facilitating and eliciting learning of the extended TPB, which integrates the TPB and among the group members. IMB, as aforementioned. To ensure the consistency and correctness of the facil- In comparison with the six-step model in quality interven- itators in conducting the intervention according to the tion development proposed by Wight and colleagues [39], intervention content and applying the communication MEDIHEALTH adopted a similar procedure for quality skills learned, three sessions of the intervention prior to control. To define and understand the problem of medica- the actual study will be observed and assessed by the re- tion adherence, a scoping review of studies investigating fac- searchers on their performance in terms of coverage of all tors affecting medication adherence among T2DM patients learning topics, consistency in conducting the intervention, Ting et al. Trials (2018) 19:310 Page 8 of 13

communication with participants and responsiveness to Table 1 Outcome measurements of the study participants’ concerns using a structured evaluation form Variables No. of items Measuring scale (Additional file 3). Besides, the facilitators will also be asked Primary outcome to evaluate their own performance on the same aspects. Medication adherence 81–7 items are measured They are required to achieve 90–100% of all the aspects [34, 35] with binomial answers, mentioned above to be qualified as facilitators for this which are “yes” or “no”; the last item is measured study. with 5-point Likert scale Each session will require one main facilitator and three Secondary outcome assistant facilitators to deliver the program. Besides, four qualified and trained facilitators are available for back- Adherence information 6 5-point Likert scale [57] up. The number of participants for each session of the Attitude to adhere 5 5-point Likert scale structured group-based intervention is fixed between 20 [58] and 24 participants. Thus, during the mini-group discus- Subjective norm to 6 5-point Likert scale sion session, each facilitator will assist an average of five adhere [58] to six participants. Notably, all facilitators are the phar- Perceived behavioural 11 5-point Likert scale macists who serve under the Pharmaceutical Services control to adhere [59] Division of Sarawak State Health Department. The set- Intention to adhere 3 5-point Likert scale ting of the venue for the Program is available in [60] Additional file 4.

Content of the control group made accordingly. The six individuals involved in the Participants who are assigned to the control group will pre-test will not be included in the actual study. be asked to complete the questionnaire with the assist- ance of the facilitators. A briefing on how to answer the questionnaire will be given by the facilitators, who can Qualitative evaluation on effectiveness and sustainability answer any questions related to the questionnaire raised On the one hand, to explore how and why the program by the participants. However, the facilitators are not impacts the medication-taking behaviours of partici- allowed to answer on behalf of the participants. After pants, a one-to-one semi-structured interview with five completion of the questionnaire, participants will be in- participants using purposive sampling will be conducted formed about the subsequent follow-up measurement 1 month after the intervention. The questions that will after 1, 3, 6 and 12 months. After that, they will be dis- be asked include (1) how would the GBEP MEDI- missed and receive their usual care provided by the HEALTH help to improve medication adherence among health clinics as before without any changes. T2DM Malay patients; (2) how did this program help you in improving your medication adherence; (3) what Outcome measurement is(are) the weakness(es) of the program and what could The primary outcome of this study is medication adher- be done to improve it; and (4) would you recommend ence, whereas the secondary outcomes are the variables this SGBI to other T2DM Malay patients and why. that contribute to the primary outcome as depicted in On the other hand, to explore sustainability in terms the conceptual framework. The details of the instrument of organizational routines and institutional standards used to measure the primary and secondary outcomes of [22], two main facilitators and two managerial officers of this study are illustrated in Table 1. Sarawak Pharmaceutical Services Division who oversee the implementation of the program will be interviewed Validation of the questionnaire after 12 months of the intervention. The aspects of sus- All items, which are originally in English, will be trans- tainability of the program that will be discussed include lated into Bahasa Malaysia and will be back-translated (1) how would the department sustain the manpower re- into English by a group of two experienced language lec- quired in implementing the program; (2) how would the turers from the Universiti Malaysia Sarawak. Two ex- department sustain the long-term implementation of the perts in the field of behavioural studies will be invited to program based on the cost involved in running the pro- examine the content validity of the BM questionnaire. gram; (3) how and why would the program gain support Furthermore, the translated questionnaire will be pre- from top management; (4) how and why would the pro- tested among six Malay patients with T2DM from PJHC gram be implemented in other facilities; and (5) how prior to the study. All comments given during the pre- and why would the program be implemented regularly. test pertinent to the questionnaire design, items and ease All interviews will be audio-recorded and transcribed of administration will be reported, and amendments verbatim for further analyses. Ting et al. Trials (2018) 19:310 Page 9 of 13

Table 2 Framework of treatment fidelity strategies Components Goal Strategies Study design Ensure the same treatment dose within 1. The structured GBEP is designed to be completed within conditions and equivalent dose across 3 h with an allowance of 15-min deviation. conditions. 2. The intervention manual will ensure all facilitators conduct the intervention in a consistent manner. 3. Observation on 3 sessions of the intervention conducted by the involved facilitators prior to the actual study will be done by the researchers to assess the consistency and appropriateness in conducting the intervention. Feedback will be given to the facilitators by the observers after the observation. The facilitators will also discuss the issues faced during the intervention with the researchers. 4. All facilitators are acquired to adhere to the time allocated for each activity throughout the intervention. Plan for implementation setbacks. Have an extra 4 qualified and trained facilitators in case of unavailability of the involved facilitators. Provider training Standardize training. All the qualified and involved facilitators together with the 4 back-up facilitators will be trained together to ensure consistency in conducting the intervention. Observation on 3 sessions of the intervention conducted by the involved facilitators prior to the actual study will be able to ensure the actual performance of the involved facilitators. Ensure provider skill acquisition. A scoring scale to assess the qualification and consistency of the facilitators in conducting the intervention will be practised. Minimize “drift” in provider skills. During the actual study, the researchers will still observe the intervention conducted by the facilitators to ensure the consistency of the intervention. Should the researchers observe below 90% of consistency as compared to the training sessions, the reasons that caused the inconsistency will be investigated and reported. Accommodate provider differences. All facilitators are pharmacists who work in the Pharmaceutical Services Division, Sarawak State Health Department. Hence, the facilitators have a similar pattern of knowledge background and are considered expert related to the study. Treatment delivery Control for provider differences. The facilitators have similar background and have the same training at the same time. Reduce differences within treatment. A scripted intervention manual is available in the form of Microsoft PowerPoint slides and used by the facilitators. Ensure adherence to the treatment During the actual study, the researchers will still observe the protocol. intervention conducted by the facilitators and will be video-recorded to ensure the consistency of the intervention. Should the researchers observe below 90% of consistency as compared to the training sessions, the reasons that caused the inconsistency will be investigated and reported. Minimize contamination between This is a randomized controlled trial with blinding on the conditions. researchers, facilitators and participants to treatment allocation prior to the intervention. Treatment receipt Ensure participant comprehension. 1. Participant understanding on the message will be evaluated with the scales developed to measure the impact of the intervention on the psychosocial variables of the participants. A comparison between the intervention group and the control group will show whether the improvement in the psychosocial variables is due to chance or is because of the intervention. 2. A qualitative interview after the intervention will enable the researchers to know how the intervention impacts their medication-taking behaviour. Ensure participant ability to use 1. The facilitators work with the participants until they can cognitive skills. demonstrate correct medication-taking skills. 2. Hypothetical situations that the participants may face in real life will be addressed during group discussion and sharing on Ting et al. Trials (2018) 19:310 Page 10 of 13

Table 2 Framework of treatment fidelity strategies (Continued) Components Goal Strategies their reasons for non-adherence and the method that they will adopt to overcome the problem. Ensure participant ability to perform The facilitators work with the participants until they can behavioural skills. demonstrate correct medication-taking skills. Enactment of treatment Ensure participant use of cognitive skills. The use of a medication chart prepared by the participants will skills show how well they comprehend the medication-taking skills. Ensure participant use of behavioural Medication adherence will be measured after 1, 3 and 6 months skills of the intervention to ensure that the messages conveyed through the intervention are translated into action and such action is maintained.

Treatment fidelity adherence, whereas a MMAS-8 score <6 will be catego- Treatment fidelity of the structured GBEP will be evalu- rized as non-adherence. ated using the concept and strategies developed by the Assumption checking will be done to decide whether a Treatment Fidelity Workgroup of the National Institutes parametric test or non-parametric test is to be employed of Health Behaviour Change Consortium [49]. The in the data analysis. The independent t test or Mann- framework of treatment fidelity strategies for this study Whitney U test will be used to examine the difference in is depicted in Table 2. medication adherence levels and psychosocial variables among participants between the intervention and control groups. To test the change in secondary outcome variables Data analysis before and after 1, 3, 6 and 12 months of the intervention, All data recorded will be scrutinized for accuracy and repeated-measures ANOVA using IBM SPSS software will completeness. Data obtained will be entered into an be used to assess the difference in the between-groups ef- IBM SPSS data sheet. Data collected will be explored, fect (inter-group comparison) and within-subject effect checked and cleaned to detect any missing data or (time effect). Repeated measures ANOVA is equivalent to wrong data entries. Data will later be entered and stored ANOVA for related samples and is the extension of the in password-protected electronic storage. Original docu- paired t test. Multivariate analysis will be used in this ments, including signed informed consent and com- study to eliminate confounding effect and prevent type 1 pleted questionnaires, will be retained by the principal error inflation. For the testing of hypotheses, path analysis investigator for a minimum of 15 years. All study docu- using a generalized estimating equation (GEE) with the ments will be stored in locked cabinets and electronic Stata Version 14 [51] xtgee command will be conducted. data encrypted with restricted access. Enumerators will The GEE procedure is an extension of the generalized lin- employ multiple attempts to follow up all the partici- ear model for analysis of repeated measurements. The pants to minimize missing outcome data. Firstly, the path coefficients will be obtained from the β weights of enumerators will check the completeness of the ques- the GEE analyses. All statistical analyses will be two-sided tionnaires upon their submission during follow-up visits. with p < 5% for consideration of statistical significance. If the participants do not show up during the follow-up For the qualitative data collected through the semi- visits, they will be contacted and interviewed via a phone structured interview, the constant comparative method call. If they cannot be contacted via a phone call, the last for qualitative analysis [52] with the assistance of resolution will be a home visit. However, should a missing ATLAS.ti 5 [53] will be used to identify the themes re- outcomedatabecomeunavoidable,multipleimputation lated to the questions. analysis using Stata MI commands with a memory-efficient style (mlong) and five imputations will be employed [50]. Discussion The overall mean will be computed using IBM SPSS While previous studies have provided little empirical evi- software for outcome variables: adherence information, dence of culturally sensitive health-promoting interven- attitude to adhere, subjective norm to adhere, perceived tions [38], only tested on specific minority populations behavioural control to adhere and intention to adhere. [54], this study aims to provide evidence on the effective- The outcome variable medication adherence will be ness of a structured GBEP that employs psychosocial and categorized into two: adherence and non-adherence, be- behavioural approaches, with multiple theoretical ground- fore the researcher proceeds with univariate and multi- ing and a culturally sensitive approach in promoting medi- variate analyses. For the outcome variable medication cation adherence among Malay patients with underlying adherence, a MMAS-8 score >6 will be categorized as T2DM. Nevertheless, the testing of the integrated model Ting et al. Trials (2018) 19:310 Page 11 of 13

(TPB and IMB) in explaining medication adherence is also Additional file 2: MEDIHEALTH content and intervention mapping. contributing to the extant body of knowledge. Moore and (PDF 193 kb) Evans argued that it is important to take the context into Additional file 3: Facilitator competency and inter-coder reliability consideration when choosing the theory that “best fits” assessment form. (PDF 300 kb) the complex population being tested [55]. Therefore, we Additional file 4: Setting of the venue for the MEDIHEALTH program. (PDF 326 kb) decided to test the two theories—TPB and IMB—among our Malay patients with T2DM. These theories have been Abbreviations applied to the testing of medication adherence and proven GBEP: Group-based educational program; HIV: Human immunodeficiency effective among patients with T2DM [24, 25]. Nonethe- virus; IMB: Information-motivation-behavioural skills model; KSHC: Kota less, the variables of the TPB and IMB were found to in- Samarahan Health Clinic; KYM: Know Your Medicine; MEDIHEALTH: Know Your Medicine – Take it for Health; MMAS-8: 8-item Morisky Medication clude the factors that influence medication adherence Adherence Scale, Malaysian specific; MOH: Ministry of Health, Malaysia; among Malay T2DM patients as revealed by the scoping DMTAC: Diabetes Medication Therapy Adherence Clinic; NHMS: National review and interview with key informants as mentioned health morbidity survey; PJHC: Petra Jaya Health Clinic; PSD: Pharmaceutical Services Division; T2DM: Type 2 diabetes mellitus; TPB: Theory of planned before. Hence, the findings of this study could enrich behaviour these existing theories through their application to a multicultural Malaysian context. Acknowledgements Firstly, we would like to thank the Director of Sarawak State Health For healthcare professionals, the quantitative and Department, Dr Jamilah Hashim; the Health Officer, Dr Azmi qualitative findings of this study could assist in identify- B Ahmad; and the Health Officer, Dr Nur Fatihah Oh ing the possible shortfall of the program and provide a Abdullah, for their approval to conduct this study. Secondly, we would like to acknowledge all the facilitators of the MEDIHEALTH program: Mr Izzul baseline for future studies. Notably, we employ both in- Syazwan B Shuib, Mdm Syahida Sarini Bt Bolhan, Mdm Siti Fairuz Bt Selbi, ductive and deductive reasoning on the mechanism of Mdm Dayangku Fazilla Bt Awang Mahrup, Mdm Nurul Ain Bt Iskandar, Ms Siti the program because we agree with Astbury and Leeuw Zakiyyah Bt Bakhtiar, Mr Ridhwan B Abdul Razak, Mr Luqman Nulhakim B “ Said and Ms Izzah Alia Bt Mohd Rosli. Also, the authors would like to that program theory building with mechanisms involves acknowledge the officers who assisted in recruiting participants: Mr James constant shuttling between theory and empirical data, Tie Lok Kin, Mr Ebson Anak Ngumbang, Mr Mohamad Faizal Bin Mansor @ using both inductive and deductive reasoning” because Din, Mr Hamidi Bin Drahman, Mdm Lily Siao, Mr Adrian Tan Lee Wee, Ms “ ” “ Rachel Sii Zu Wen, Mr Elwin Jong Chai Ming, Mr Brian A/L Navaratnam, Ms the mechanisms are usually hidden and sensitive to Rodgne Alexius Anak Sultan, Mr Mohamad Khirul Anuar Bin Mohd Noor and variations in context” [56]. Mr Syazwan Bin Manshor. We also thank the family physician and chief Lastly, the findings on the effectiveness of the GBEP in pharmacist of PJHC, Dr Juslina Omar and Mdm Esther Ngau, respectively, and the family physician and chief pharmacist of KSHC, Dr Muhamad Irfan improving medication adherence could become evidence Yasin Bin Ali Uddin and Ms Siti Noraisah Bt Kifli, respectively. We thank the for policymakers and authorities in justifying the resources University of Oxford and University of Cambridge for permission to translate spent in running the program. Furthermore, this one-off, and use the scale to measure the belief about taking hypoglycaemic medication among people with T2DM. We thank Professor Dr Mary Lynn pharmacist-led and multiple-theoretical-grounding struc- McPherson, from the University of Maryland School of Pharmacy, and her tured GBEP module could be expanded to other types of colleagues for their permission to translate and to use the scale to measure medication management, which could complement the knowledge about medication among diabetic patients. We thank Professor Dr Gbenga Ogedegbe, from the New York University School of Medicine, existing DMTAC in improving medication adherence. It is and his colleagues for their permission to use the scale to measure expected that the combination of both individual and medication adherence self-efficacy. We thank Dr Aaron T. Vissman for his per- group-based approaches could maximize the coverage of mission to use and translate the scale to measure intention to adhere to medications. We also thank Mdm Monaliza Bt Hj. Sarbini and Mdm Rosnah patients who require special care in their medication man- Bt Mustafa, from Universiti Malaysia Sarawak Faculty of Language and agement and reduce the disease burden due to medication Communication, for their contributions to translating the questionnaire from non-adherence. English to Malay and back-translating from Malay to English. The MMAS-8 content, name and trademarks are protected by US copyright and trademark laws. Permission to use the scale and its coding is required. A license agreement is available from Donald E. Morisky, ScD, ScM, MSPH, 14725 NE Trial status 20th St Bellevue, WA 98007, USA ([email protected]). This protocol is registered with ClinicalTrials.gov with trial Funding identifier NCT03228706. It was first posted on July 25, This study is funded by Sarawak Research Society for the use of the MMAS-8. 2017, and was updated on January 25, 2018 (version 1.3; The funder had no role in the study design, data collection and analysis, or https://clinicaltrials.gov/ct2/show/NCT03228706). Recruit- publication. ment began on August 1, 2017, and is expected to be Availability of data and materials completed on July 31, 2018. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Authors’ contributions Additional files CYT conceived of the study, designed the intervention, prepared the study protocol and contributed to the writing of the manuscript. SAZD and MAH prepared the study protocol and contributed to the writing of the manuscript. Additional file 1: SPIRIT fillable checklist. (PDF 179 kb) HT prepared the study protocol and the data analysis plan. ISS designed the Ting et al. Trials (2018) 19:310 Page 12 of 13

intervention of this study and trained the facilitators. SCL, AHAAJ, NAO, SEL and 9. Mohamad Azmi H, Fahad S, Mohd Dziehan M, Che Pun B, Salmiah MA, Siti STS designed the intervention of this study and contributed to the writing of Fauziah A, Norazlin AK, Abdul Haniff MY, Kamarudin A, Siew LJ, Lai ST. A the manuscript. LCJ prepared the data analysis plan and contributed to the national survey on the use of medicines (NSUM) by Malaysian consumers. writing of the manuscript. RT and DEM contributed to the writing of the Pharmaceutical Services Division, Ministry of Health Malaysia. 2016. https:// manuscript. All authors read and approved the final manuscript. www.pharmacy.gov.my/v2/sites/default/files/document-upload/national- survey-use-medicine-iii-nsum-iii-1.pdf. Accessed 15 Mar 2017. Ethics approval and consent to participate 10. Capoccia K, Odegard PS, Letassy N. Medication adherence with diabetes All data are restricted to the principal investigator and solely used for research medication: a systematic review of the literature. Diabetes Educ. 2016; purposes. The study will be conducted in compliance with the ethical 42(1):34–71. principles outlined in the Declaration of Helsinki and Malaysian Good Clinical 11. American Diabetes Association. Standards of medical care in diabetes: 2013. Practice Guideline. Written informed consent from all participants will be Diabetes Care. 2013;36 Suppl 1:S11–66. obtained when they agree to participate in this study. This trial is registered 12. Meng J, Casciano R, Lee YC, Stern L, Gultyaev D, Tong L, Kitio-Dschassi with the Medical Research and Ethics Committee (MREC), National Institutes of B. Effect of diabetes treatment-related attributes on costs to type 2 Health Malaysia, prior to the actual study with approval number (12)KKM/ diabetes patients in a real-world population. J Manag Care Spec Pharm. NIHSEC/P17-854. The trial is registered with ClinicalTrials.gov following approval 2017;23(4):446–52. from the MREC with trial registration number NCT03228706. 13. Pharmaceutical Services Division, Ministry of Health Malaysia. Know Your Medicine. 2017. http://www.knowyourmedicine.gov.my. Accessed Competing interests 10 Oct 2017. Morisky receives an honorarium for the use of the MMAS-8 and was not 14. Sarawak State Health Department. MEDIHEALTH program. 2018. http:// involved in the data analysis. All other authors declare that they have no jknsarawak.moh.gov.my/bm/modules/mastop_publish/?tac=133. Accessed competing interests. The principal investigator will take the responsibility for 28 Jan 2017. reporting any competing interest that emerges during the study. 15. Pharmaceutical Services Division, Ministry of Health Malaysia. Perkhidmatan Medication Therapy Adherence Clinic (MTAC). 2013. https://www.pharmacy. gov.my/v2/ms/entri/perkhidmatan-medication-therapy-adherance-clinic- Publisher’sNote mtac.html. Accessed 20 Mar 2017. Springer Nature remains neutral with regard to jurisdictional claims in 16. Mattson M, Hall JG. Linking health communication with social support. published maps and institutional affiliations. Junker Shutter Stock. 2011;23(4):22–45. 17. Davies MJ, Heller S, Skinner TC, Campbell MJ, Carey ME, Cradock S, Dallosso Author details HM, Daly H, Doherty Y, Eaton S, Fox C. Effectiveness of the diabetes 1Institute of Borneo Studies, Universiti Malaysia Sarawak, Kota Samarahan, education and self- management for ongoing and newly diagnosed Sarawak, Malaysia. 2Discipline of Social and Administrative Pharmacy, School (DESMOND) programme for people with newly diagnosed type 2 diabetes: of Pharmaceutical Sciences, Universiti Sains Malaysia, Penang, Malaysia. cluster randomised controlled trial. BMJ. 2008;336(7642):491–5. 3Sarawak Research Society, Sarawak, Malaysia. 4Clinical Research Center, 18. Odgers-Jewell K, Ball LE, Kelly JT, Isenring EA, Reidlinger DP, Thomas R. Sarawak General Hospital, Sarawak, Malaysia. 5Jeffrey Cheah School of Effectiveness of group-based self-management education for individuals Medicine and Health Sciences, Monash University Malaysia, Selangor, with type 2 diabetes: a systematic review with meta-analyses and meta- Malaysia. 6Pharmaceutical Services Division, Sarawak State Health regression. Diabet Med. 2017;34(8):1027–39. Department, Sarawak, Malaysia. 7Pharmacy Practice and Development 19. Edmondson D, Falzon L, Sundquist KJ, Julian J, Meli L, Sumner JA, Kronish Division, Sarawak State Health Department, Sarawak, Malaysia. 8Pharmacy IM. A systematic review of the inclusion of mechanisms of action in NIH- Enforcement Division, Sarawak State Health Department, Sarawak, Malaysia. funded intervention trials to improve medication adherence. Behav Res 9Department of Community Health Sciences, UCLA Fielding School of Public Ther. 2018;101:12–9. Health, Los Angeles, CA, USA. 20. O'Loughlin J, Renaud L, Richard L, Gomez LS, Paradis G. Correlates of the sustainability of community-based heart health promotion interventions. Received: 31 January 2018 Accepted: 13 April 2018 Prev Med. 1998;27(5):702–12. 21. Shediac-Rizkallah MC, Bone LR. Planning for the sustainability of community- based health programs: conceptual frameworks and future directions for – References research, practice and policy. Health Educ Res. 1998;13(1):87 108. 1. World Health Organization. Global report on diabetes: World Health 22. Pluye P, Potvin L, Denis JL. Making public health programs last: – Organization; 2016. http://apps.who.int/iris/bitstream/10665/204871/1/ conceptualizing sustainability. Eval Program Plann. 2004;27(2):121 33. 9789241565257_eng.pdf. Accessed 18 Mar 2017 23. Stirman SW, Kimberly J, Cook N, Calloway A, Castro F, Charns M. The sustainability 2. Kanapathy J. Illness perception of type 2 diabetic patients in Malaysia of new programs and innovations: a review of the empirical literature and (Doctoral dissertation, City University London). 2015. recommendations for future research. Implement Sci. 2012;7(1):17. 3. Institute for Public Health. Vol. II: non-communicable diseases, risk factors & 24. Rich A, Brandes K, Mullan B, Hagger MS. Theory of planned behavior and other health problems. In: National Health and Morbidity Survey 2015. adherence in chronic illness: a meta-analysis. J Behav Med. 2015;38(4):673–88. National Institutes of Health, Ministry of Health Malaysia; 2015. http://iku. 25. Mayberry LS, Osborn CY. Empirical validation of the information– moh.gov.my/images/IKU/Document/REPORT/nhmsreport2015vol2.pdf. motivation–behavioral skills model of diabetes medication adherence: a Accessed 18 Mar 2017. framework for intervention. Diabetes Care. 2014;37(5):1246–53. 4. Samiei V, Puteh SE, Manaf MR, Latip KA, Ismail A. Are Malaysian diabetic 26. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process. patients ready to use the new generation of health care service delivery? A 1991;50(2):179–211. telehealth interest assessment. Malays J Med Sci. 2016;23(2):44. 27. Bandura A. Self-efficacy: toward a unifying theory of behavioral change. 5. Department of Statistics Malaysia. Population and housing census of Psychol Rev. 1977;84(2):191. Malaysia 2010: Sarawak. 2010. https://web.archive.org/web/ 28. Munro S, Lewin S, Swart T, Volmink J. A review of health behaviour theories: 20141114152728/http://www.statistics.gov.my/portal/download_ how useful are these for developing interventions to promote long-term Population/files/population/04Jadual_PBT_negeri/PBT_Sarawak.pdf. medication adherence for TB and HIV/AIDS? BMC Public Health. 2007;7(1):104. Accessed 15 Apr 2017. 29. Fisher JD, Fisher WA, Amico KR, Harman JJ. An information-motivation- 6. Krass I, Schieback P, Dhippayom T. Adherence to diabetes medication: a behavioral skills model of adherence to antiretroviral therapy. Health systematic review. Diabet Med. 2015;32(6):725–37. Psychol. 2006;25(4):462. 7. García-Pérez LE, Álvarez M, Dilla T, Gil-Guillén V, Orozco-Beltrán D. 30. Butt M, Ali AM, Bakry MM, Mustafa N. Impact of a pharmacist led diabetes Adherence to therapies in patients with type 2 diabetes. Diabetes Ther. mellitus intervention on HbA1c, medication adherence and quality of life: a 2013;4(2):175–94. randomised controlled study. Saudi Pharm J. 2016;24(1):40–8. 8. World Health Organization. Adherence to long-term therapies. Evidence for 31. Dupont WD, Plummer WD. Power and sample size calculations for studies action. Geneva: World Health Organization; 2003. involving linear regression. Control Clin Trials. 1998;19(6):589–601. Ting et al. Trials (2018) 19:310 Page 13 of 13

32. Sabaté E. Adherence to long-term therapies: evidence for action. Geneva: 57. McPherson ML, Smith SW, Powers A, Zuckerman IH. Association between World Health Organization; 2003. diabetes patients’ knowledge about medications and their blood glucose 33. Kirkman MS, Rowan-Martin MT, Levin R, et al. Determinants of adherence to control. Res Soc Adm Pharm. 2008;4(1):37–45. diabetes medications: findings from a large pharmacy claims database. 58. Farmer A, Kinmonth AL, Sutton S. Measuring beliefs about taking hypoglycaemic Diabetes Care. 2015;38(4):604–9. https://doi.org/10.2337/dc14-2098. medication among people with type 2 diabetes. Diabet Med. 2006;23(3):265–70. 34. Morisky DE, Ang A, Krousel-Wood M, Ward H. Predictive validity of a 59. Fernandez S, Chaplin W, Schoenthaler AM, Ogedegbe G. Revision and medication adherence measure for hypertension control. J Clin Hyperten. validation of the Medication Adherence Self-Efficacy Scale (MASES) in 2008;10(5):348–54. hypertensive African Americans. J Behav Med. 2008;31(6):453–62. 35. Chung WW, Chua SS, Lai PS, Morisky DE. The Malaysian Medication 60. Vissman AT, Young AM, Wilkin AM, Rhodes SD. Correlates of HAART Adherence Scale (MALMAS): concurrent validity using a clinical adherence among immigrant Latinos in the southeastern United States. measure among people with type 2 diabetes in Malaysia. PLoS One. AIDS Care. 2013;25(3):356–63. 2015;10(4):e0124275. 36. Suresh KP. An overview of randomization techniques: an unbiased assessment of outcome in clinical research. J Hum Reprod Sci. 2011;4(1):8. 37. Slater MD. Integrating application of media effects, persuasion, and behavior change theories to communication campaigns: a stages-of-change framework. Health Commun. 1999;11(4):335–54. 38. Mier N, Ory MG, Medina AA. Anatomy of culturally sensitive interventions promoting nutrition and exercise in Hispanics: a critical examination of existing literature. Health Promot Pract. 2010;11(4):541–54. 39. Wight D, Wimbush E, Jepson R, Doi L. Six steps in quality intervention development (6SQuID). J Epidemiol Community Health. 2015; https://doi. org/10.1136/jech-2015-205952. 40. Neoh CF, Long CM, Lim SM, Ramasamy K, Shahar S, Majeed AB. Medication use and adherence among multi-ethnic community-dwelling older adults in Malaysia. Geriatr Gerontol Int. 2017;17(8):1214–20. 41. Omar MS, San KL. Diabetes knowledge and medication adherence among geriatric patient with type 2 diabetes mellitus. Malay. 2014;36:53–4. 42. Al-Qazaz HK, Sulaiman SA, Hassali MA, Shafie AA, Sundram S, Al-Nuri R, Saleem F. Diabetes knowledge, medication adherence and glycemic control among patients with type 2 diabetes. Int J Clin Pharm. 2011;33(6):1028–35. 43. Manan MM, Husin AR, Alkhoshaiban AS, Al-Worafi YM, Ming LC. Interplay between oral hypoglycemic medication adherence and quality of life among elderly type 2 diabetes mellitus patients. J Clin Diagn Res. 2014;8(12):JC05. 44. Junaini SN, Hwey AL, Sidi J, Rahman KA. Development of Sarawak Malay Local Dialect Online Translation Tool. In Computer Technology and Development. ICCTD'09. International Conference on 2009 Nov 13, Vol 1. IEEE; 2009. pp. 459–62. 45. Soto-Greene ML, Salas-Lopez D, Sanchez J, Like RC. Antecedents to effective treatment of hypertension in Hispanic populations. Clinical Cornerstone. 2004;6(3):30–8. 46. Jones S. Putting the person into person-centered and immediate emotional support emotional change and perceived helper competence as outcomes of comforting in helping situations. Commun Res. 2004;31(3):338–60. 47. Anderson RM, Funnell MM, Butler PM, Arnold MS, Fitzgerald JT, Feste CC. Patient empowerment. Results of a randomized controlled trial. Diabetes Care. 1995;18:943–9. 48. Anderson RM. Patient empowerment and the traditional medical model. A case of irreconcilable differences? Diabetes Care. 1995;18:412–5. 49. Bellg AJ, Borrelli B, Resnick B, Hecht J, Minicucci DS, Ory M, Ogedegbe G, Orwig D, Ernst D, Czajkowski S. Enhancing treatment fidelity in health behavior change studies: best practices and recommendations from the NIH Behavior Change Consortium. Health Psychol. 2004;23(5):443. 50. StataCorp. Stata 14 base reference manual. College Station, TX: Stata Press; 2015. 51. StataCorp. Stata 13 base reference manual. College Station, TX: Stata Press; 2013. 52. Glaser BG. The constant comparative method of qualitative analysis. Soc Probl. 1965;12(4):436–45. 53. Muhr T, Friese S. User’s manual for ATLAS.ti 5.0. Berlin: ATLAS.ti Scientific Software Development GmbH; 2004. 54. Hodge DR, Jackson KF, Vaughn MG. Culturally sensitive interventions and health and behavioral health youth outcomes: a meta-analytic review. Soc Work Health Care. 2010;49(5):401–23. 55. Moore GF, Evans RE. What theory, for whom and in which context? Reflections on the application of theory in the development and evaluation of complex population health interventions. SSM-Population Health. 2017;3:132–5. 56. Astbury B, Leeuw FL. Unpacking black boxes: mechanisms and theory building in evaluation. Am J Eval. 2010;31(3):363–81.